Even when there is initial therapeutic sensitivity to a conventional chemotherapy or targeted therapy, tumors can become resistant and recur. Methods that model and predict therapeutic resistance of cancer can be extremely useful in the development of more effective treatments for cancer. Our long-term goal is to develop strategies to model, predict, and target therapeutic resistance of cancer. Our proposed approach is to utilize conditional transgenic mouse models combined with computational modeling. We hypothesize that therapeutic resistance to oncogene inactivation can be modeled and thus predicted as a consequence of clonal evolution of tumor cells driven by both cell autonomous and immune-mediated selective pressures. Our approach in Aim 1 is to build a mathematical model that incorporates the roles of the immune system and of evolutionary dynamics to predict the emergence of therapeutic resistance upon oncogene inactivation. Then, in Aim 2 we will experimentally interrogate the roles of the immune system and of evolutionary dynamics in the emergence of therapeutic resistance. We will directly examine immune effectors/cytokines and clonal evolution in our conditional transgenic mouse model with intravital microscopy and bioluminescence imaging. Finally, in Aim 3 we will validate in vivo our mathematical model's predictions of the emergence of therapeutic resistance under novel circumstances.